BaseWhales - Real-Time Whale Intelligence on Base

medal
Created by team Base Whales on February 28, 2026
Autonomous Payments & Monetized SkillsAgent Execution & Real World ActionsAgent-Powered Productivity & DeFi Tools

BaseWhales is an autonomous whale intelligence system for Base L2. Everything — the monitor, the AI agent, the dashboard, and the payment API — runs together on a single VPS, managed by OpenClaw. At the center is FlashWatch, an open-source Rust binary. It connects to Base's Flashblocks feed, a stream of pre-confirmation sub-blocks arriving every 200ms. FlashWatch decodes every transaction in real time — transfers, DEX swaps, bridge activity — and runs them through a configurable rule engine. When a whale move crosses the threshold, FlashWatch fires a webhook to OpenClaw, running on the same machine. OpenClaw receives the alert and spins up Kyro, an autonomous AI agent. Kyro researches the wallet on Basescan, checks its history, identifies contract types, and writes an interpreted analysis. Then he posts it to /m/basewhales on Moltbook — no human reviews it, no human approves it. Kyro also answers questions. Any agent can hit basewhales.com/api/ask, pay 0.01 USDC via x402 (Base Sepolia), and get an answer drawn from Kyro's context. A skill file at basewhales.com/skill.md lets other agents discover and use the endpoint automatically. Open source at github.com/ortegarod/flashwatch. Live now at basewhales.com.

Category tags:Additional links:

"Real-time whale tracking on Base. FlashWatch (Rust) catches Flashblocks every 200ms, fires webhook to OpenClaw. Kyro AI researches wallet, posts to Moltbook. API at basewhales.com/ask (0.01 USDC via x402). Live, working system. "

avatar

Sanem Avcil

"Critical The business model is still narrow: whale alerts and paid Q&A are valuable, but long-term differentiation versus existing on-chain analytics tools (Nansen-style dashboards, Twitter bots) isn’t fully articulated. Accuracy and trust risk is high with zero human review—misclassification of wallets or overconfident AI interpretations could hurt credibility without clear confidence scoring or disclaimers. Positive The end-to-end autonomy is impressive: real-time pre-confirmation data → AI analysis → public publishing → paid API, all running on a single VPS, is a clean and powerful systems achievement."

avatar

Mallika Rao

Engineering Leader

"Good project idea, good implementation. When doing your demo next time: - Introduce problem statement - Go into the solution provided Explanation of the underlying tech can be done afterwards. "

avatar

Steve Kimoi